Reproductive Outcomes Following Induced Abortion: a National Register- Based Cohort Study in Scotland

Open Access Research BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from Reproductive outcomes following induced abortion: a national register- based cohort study in Scotland

Siladitya Bhattacharya,1 Alison Lowit,1 Sohinee Bhattacharya,1 Edwin Amalraj Raja,1 Amanda Jane Lee,1 Tahir Mahmood,2 Allan Templeton1

To cite: Bhattacharya S, ABSTRACT ARTICLE SUMMARY Lowit A, Bhattacharya S, Objective: To investigate reproductive outcomes in et al . Reproductive outcomes women following induced abortion (IA). Article focus following induced abortion: Design: Retrospective cohort study. - Is an IA in a first pregnancy associated with a national register-based cohort study in Scotland. Setting: Hospital admissions between 1981 and 2007 spontaneous preterm birth or other adverse BMJ Open 2012;2:e000911. in Scotland. obstetric or perinatal outcomes in the second doi:10.1136/ Participants: Data were extracted on all women who pregnancy? bmjopen-2012-000911 had an IA, a miscarriage or a live birth from the - Is an IA performed after an initial singleton live Scottish Morbidity Records. A total of 120 033, birth associated with spontaneous preterm birth < Prepublication history and 457 477 and 47 355 women with a documented or adverse obstetric or perinatal outcomes in the an additional table for this next pregnancy? paper are available online. To second pregnancy following an IA, live birth and - Do any of these associations differ by method of view these files please visit miscarriage, respectively, were identified. the journal online Outcomes: Obstetric and perinatal outcomes, especially IA (ie, surgical vs medical)? (http://dx.doi.org/10.1136/ preterm delivery in a second ongoing pregnancy - Is the risk of adverse obstetric or perinatal bmjopen-2012-000911). following an IA, were compared with those in outcomes associated with increasing number of primigravidae, as well as those who had a miscarriage terminations? Received 20 January 2012 or live birth in their first pregnancy. Outcomes after Key messages Accepted 27 June 2012 surgical and medical termination as well as after one or - The risk of preterm birth after IA is lower more consecutive IAs were compared. This final article is available than that after miscarriage but higher than http://bmjopen.bmj.com/ for use under the terms of Results: IA in a first pregnancy increased the risk of that in a first pregnancy or after a previous live the Creative Commons spontaneous preterm birth compared with that in birth. Attribution Non-Commercial primigravidae (adjusted RR (adj. RR) 1.37, 95% - This risk is not increased further in women who 2.0 Licence; see CI 1.32 to 1.42) or women with an initial live birth undergo two or more consecutive IAs. http://bmjopen.bmj.com (adj. RR 1.66, 95% CI 1.58 to 1.74) but not in - Surgical but not medical abortion appears to be comparison with women with a previous miscarriage associated with an increased risk of spontaneous (adj. RR 0.85, 95% CI 0.79 to 0.91). Surgical abortion preterm birth in comparison with primigravidae. increased the risk of spontaneous preterm birth

compared with medical abortion (adj. RR 1.25, 95% CI on September 24, 2021 by guest. Protected copyright. 1.07 to 1.45). The adjusted RRs (95% CI) for 2009, 13 005 abortions were performed in spontaneous preterm delivery following two, three and Scotland, with the highest rates in women four consecutive IAs were 0.94 (0.81 to 1.10), 1.06 aged 16e19 years.1 What is not yet entirely (0.76 to 1.47) and 0.92 (0.53 to 1.61), respectively. clear is the effect these abortions may have Conclusions: The risk of preterm birth after IA is on subsequent childbearing. It has been lower than that after miscarriage but higher than that in believed that infection, cervical trauma and a ﬁrst pregnancy or after a previous live birth. This risk endometrial curettage associated with is not increased further in women who undergo two or more consecutive IAs. Surgical abortion appears to be induced abortion (IA) could lead to future associated with an increased risk of spontaneous infertility, ectopic pregnancy, preterm preterm birth in comparison with medical termination delivery and placenta praevia, but the data 1Division of Applied Health of pregnancy. Medical termination was not associated from existing observational studies are Sciences, University of 2e18 Aberdeen, Aberdeen, UK with an increased risk of preterm delivery compared to mixed. Following the legalisation of 2Department of Obstetrics primigravidae. abortion in 1967, initial research on the and Gynaecology, Victoria effects of an IA on subsequent pregnancies Hospital, Kirkcaldy, UK showed no evidence of an increased risk of miscarriage, preterm delivery or low birth Correspondence to BACKGROUND 19 20 Dr Sohinee Bhattacharya; weight. Much of the work in the subject sohinee.bhattacharya@abdn. Many women start their reproductive careers has been hampered by methodological limi- ac.uk with an abortion in their ﬁrst pregnancy. In tations; randomised controlled studies are

Bhattacharya S, Lowit A, Bhattacharya S, et al. BMJ Open 2012;2:e000911. doi:10.1136/bmjopen-2012-000911 1 Reproductive outcomes following induced abortion BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from

The Scottish Morbidity Record (SMR) system in Scot- ARTICLE SUMMARY land covers a national population and has captured data Strengths and limitations of this study on medical and surgical abortion for many years. Over - Largest population-based study of reproductive outcomes 99.3% of abortions in Scotland are carried out in NHS following an IA generalisable to other populations with similar premises and are recorded in the SMR system. As these healthcare system. data are based on clinical records, any potential bias - Compares outcomes after medical and surgical abortion and created by under-reporting will be removed. The avail- explores the dose-dependent effect of abortion on future ability of this large national data set provides an ideal preterm delivery. An added strength is use of national data and opportunity to link records on abortion (SMR01) with the ability to discriminate between spontaneous and overall maternity records (SMR02) in order to explore the risk preterm birth as an outcome. - In acknowledgement of changes in clinical practice during the of preterm delivery and other maternal and perinatal long study period, models are adjusted for year of pregnancy. outcomes in women following one or more episodes of - Compares women with IA with those with a miscarriage, live IA. The data would also allow a meaningful comparison birth and nulliparous women groups, adding validity to the of outcomes following alternative forms of IA (ie, results. medical vs surgical). - Unrecorded and missing data in relation to certain potential The primary aim of this study was to investigate the confounding factors within the data set. reproductive outcomes in women following IA. In - Parity number was less reliable in the early years of data particular, we wished to answer the following research collection. This may reflect problems with coding and could questions: (1) Is an IA in a first pregnancy associated with potentially affect the quality of our results. spontaneous preterm birth or other adverse obstetric or - In addition, the analysis of such a large population-based data perinatal outcomes in the second pregnancy? (2) Is an set has the capacity to produce statistically significant differences, which may or may not be clinically relevant, IA performed after a singleton term first pregnancy associ- although this has been minimised by our use of a 1% ated with spontaneous preterm birth or adverse obstetric significance level throughout. or perinatal outcomes in the next pregnancy? (3) Do any of these associations differ by method of IA (ie, surgical vs medical)? (4) Is the risk of adverse obstetric or peri- not feasible in this context and researchers have looked natal outcomes associated with increasing number of to observational studies. Many of the published studies terminations? have been limited by small sample sizes, self-reported outcomes and inability to adjust for many potential confounders. A recent review21 reported that half of the METHODS

12 relevant studies found an association between IA and A retrospective cohort study design was used on http://bmjopen.bmj.com/ preterm birth as well as placenta praevia. More recently, routinely collected data extracted from the ISD data- a number of large studies found no increased risk of base. Approval was obtained from the Privacy Advisory placenta praevia but supported an association with Committee of the NHS, Scotland. preterm18 22 23 and very preterm delivery24 25 The clin- Data were extracted from the ISD databases (SMR01 ical implications of this are profound as reducing the and SMR02) on women aged 15e55 years who had an incidence of preterm delivery, with its considerable IA, a miscarriage, a live birth or an ongoing pregnancy associated problems, remains one of the most signiﬁcant and live delivery in their ﬁrst pregnancy between 1981

challenges in obstetrics. and 2007, followed by a second pregnancy event. on September 24, 2021 by guest. Protected copyright. Over a quarter of IAs in Scotland in 2005 were repeat Reproductive outcomes in the subsequent pregnancy of procedures1 (Information and Statistics Division (ISD), women who had an IA in their first pregnancy (exposed personal communication). While the reproductive cohort) were compared with those in two unexposed sequelae of repeat abortions are unclear, the available groups: (1) women in their second pregnancy after literature suggests that the risk of preterm delivery is a miscarriage in their first pregnancy and (2) women in increased by multiple abortions.18 22 24 26 their second pregnancy after a live birth in their first Changes in the technique of abortion have to be taken pregnancy. In addition to these two unexposed cohorts, into account when assessing their impact on future obstetric and perinatal outcomes in the subsequent reproduction. In 1992, 83.6% of terminations were pregnancy of women who had an IA in their first preg- carried out surgically, falling to 60.6% in 1998 and 40.7% nancy (exposed group) were also compared with those in 2006, with the remainder being carried out medi- women in their first pregnancy. e cally.1 A number of studies27 29 have compared these To explore outcomes following early pregnancy loss methods in terms of safety, efficacy and short-term after an initial live birth, data were extracted on all complications, but data on subsequent reproductive women (aged 15e55 years) who had an IA, a miscarriage outcomes are scant. A recent study30 found no differ- or a live birth in their second pregnancy (following a live ence in reproductive outcomes (ectopic pregnancy, birth in their first pregnancy) between 1981 and 2007 miscarriage and preterm delivery) following medical and from the ISD databases (SMR01 and SMR02) and surgical IAs but was unable to adjust for known followed up to identify a third pregnancy event. Repro- confounders, such as smoking. ductive, obstetric and perinatal outcomes in women who

2 Bhattacharya S, Lowit A, Bhattacharya S, et al. BMJ Open 2012;2:e000911. doi:10.1136/bmjopen-2012-000911 Reproductive outcomes following induced abortion BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from had an IA after a singleton term first pregnancy in the chances of a preterm birth (ie, an OR of 1.09), (exposed group) were compared with those in two assuming that the prevalence of live births in the unexposed groups: (1) women in their third pregnancy unexposed group was 6%. following a singleton term delivery in the first pregnancy and a miscarriage in the second pregnancy Statistical analysis and (2) women in their third pregnancy following two In the absence of an ideal comparison group for women singleton term deliveries. with a prior abortion, we used three unexposed cohorts, Women treated by different methods of IA (surgical or which could increase the chance of false-positive associ- medical) in their first pregnancy were compared in ations (type I error). To help minimise this, we terms of reproductive, obstetric and perinatal outcomes. used a stringent p value of #0.01 to denote statistical Finally, to answer research question 4, reproductive and significance throughout the statistical analyses. perinatal outcomes were compared between women who A generalised linear model was used with Poisson had one, two, three and four previous consecutive IAs family and robust variance estimator to ascertain the and women with no previous abortions. Each group of relationship between exposure (first pregnancy IA) and women was independent of the others, for example, various reproductive outcomes (stillbirth, miscarriage, women who had three abortions were excluded from the ectopic pregnancy and IA), maternal and perinatal group with two abortions. For each analysis, except outcomes (pre-eclampsia, placenta praevia, placental research question 4, the women were matched on parity abruption) after adjusting for potential confounders as the risk of adverse obstetric outcomes is dependent on (maternal age, year of delivery, smoking and Carstairs at parity, with primiparous women having the highest risk. relevant pregnancy). For the outcome of induction of labour, pre-eclampsia, placenta praevia and placental Data extracted abruption were also entered into the model. Similarly, The following variables were identified by matching the outcome low birth weight was also adjusted for SMR01 and SMR02 data sets between the years 1981 gestational age. Stata V.11 was used for the analysis and and 2007. a stringent p value of #0.01 was used to denote statistical Demographic details: Age at pregnancy events, smoking significance throughout. status and social class (assessed using Carstairs category As smoking data were not routinely collected in the of deprivation) in the exposed group were compared maternity database (SMR02) before 1992 and rarely with each of the three unexposed cohorts. recorded for women having an IA or miscarriage. Thus, IA details: Estimated gestation and method of termi- self-reported smoking status, collected at antenatal nation (medical or surgical or both) were recorded for booking visit, though available for some women, was the exposed group. non-randomly missing for a high percentage of women. http://bmjopen.bmj.com/ Reproductive outcomes: Miscarriage, abortion, live birth, This sometimes led to non-convergence of the statistical ectopic pregnancy, stillbirth in the exposed group were models. Therefore, a sensitivity analysis was carried out compared with the unexposed cohorts. by rerunning all the multivariate models, excluding the Obstetric and perinatal outcomes: The incidence of pre- smoking variable to determine if the overall effect sizes eclampsia, placenta praevia, placental abruption, remained of similar magnitude. This was found to be so. preterm delivery, very preterm delivery, low birth weight and the mode of delivery in the exposed cohort were RESULTS compared with each of the three unexposed cohorts. Demographic characteristics of women who had an on September 24, 2021 by guest. Protected copyright. Spontaneous delivery rates (including live birth and abortion in their first pregnancy were compared with stillbirth) were calculated after excluding women who those who had either a live birth or a miscarriage in their had induced labour and elective (planned) caesarean first pregnancy and with primigravida women (table 1). section. Women with a previous IA were significantly older, more Socioeconomic status was assessed using the Carstairs socially deprived and more likely to be smokers than Index,31 which was divided into quintiles for analysis. primigravida women or those who had a live birth or a miscarriage in a previous pregnancy. Power calculation Table 2 presents reproductive outcomes in a subse- Given the number of subgroups in the analysis coupled quent pregnancy following IA, live birth and miscarriage with multiple outcomes, a global sample size calculation in the first pregnancy. As table 2 shows, women with an was not feasible. Preliminary enquiries with ISD IA in the first pregnancy were more at risk of having suggested that we could identify at least 260 000 termi- a stillbirth or an IA in the second pregnancy compared nations (1981e2007), of which 30% (n¼69 000) were with an initial live birth. Compared with those who had estimated to have had a subsequent live birth and 25.5% an initial miscarriage, women who had an IA in their first (n¼66 223) were IAs in a first pregnancy. pregnancy were less likely to have a subsequent miscar- Using a 1:1 ratio of women with IAs in a first preg- riage or ectopic pregnancy, but more likely to have nancy (exposed cohort) and unexposed women, we another IA. anticipated having over 90% power, at the two-sided 5% Perinatal outcomes in the next ongoing pregnancy significance level, to detect a difference of 0.5% or more following IA are also compared with those in

Bhattacharya S, Lowit A, Bhattacharya S, et al. BMJ Open 2012;2:e000911. doi:10.1136/bmjopen-2012-000911 3 Reproductive outcomes following induced abortion BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from

Table 1 Demographic characteristics at first pregnancy of women who had induced abortion, live birth or miscarriage in their first pregnancy Outcome in first pregnancy Induced abortion Live birth Miscarriage (N[120 033) (N[457 477) p Value (N[47 355) p Value Mean age (SD) 24.68 (7.56) 24.89 (5.11) <0.001 26.26 (6.13) <0.001 Carstairs category* y 1 17 265 (17.1) 79 705 (18.0) <0.001 8403 (18.8) <0.001 2 18 538 (18.3) 81 661 (18.4) 8206 (18.4) 3 19 530 (19.3) 84 559 (19.1) 8794 (19.7) 4 21 135 (20.9) 92 504 (20.9) 9426 (21.1) 5 24 615 (24.4) 105 313 (23.7) 9788 (21.9) Smoking statusy Never 1014 (42.3) 112 744 (48.4) <0.001 4892 (39.8) <0.001 Current 676 (28.2) 72 182 (31.0) 2044 (16.6) Former 85 (3.5) 22 140 (9.5) 533 (4.3) Not known 622 (26.0) 26 088 (11.2) 4818 (39.2) Total 2397 233 154 12 287 Missing 117 636 (98.0) 224 323 (49.0) 35 068 (74.1) Interpregnancy interval in weeks Median (IQR) 165 (78e321) 139 (95e213) <0.001 65 (47e104) <0.001 Values are n (%) unless otherwise specified. *Carstairs categories: 1¼ least deprived, 5¼ most deprived. yPercentage based on available information for each group.

primigravida and women who have had a live birth or live birth. The risk of miscarriage in a third pregnancy miscarriage ( table 2). Compared with women having was lower in women who had an IA in their second a previous live birth, women who had an IA were at pregnancy, but the risks of another IA were higher than higher risk of pre-eclampsia; abruptio placenta; induc- in women with a previous miscarriage. tion of labour; spontaneous preterm, very preterm Compared with women with two previous live births, < < ( 32 weeks) and extremely preterm ( 28 weeks) women with a live birth followed by an IA were more http://bmjopen.bmj.com/ delivery and delivery of a low birthweight baby (<2500 g) likely to have pre-eclampsia, placenta praevia, induced but not placenta praevia. labour, low birth weight and spontaneous preterm, very In comparison with women with a previous miscar- preterm and extremely preterm birth ( table 4). Women riage, a history of IA was associated with a lower risk of with an IA in their second pregnancy were not at any developing pre-eclampsia and spontaneous preterm and signiﬁcantly higher risk of perinatal complications in very preterm delivery. Risks of pre-eclampsia, placental comparison with women with a previous miscarriage. abruption (but not placenta praevia), delivery of a low In records where the method of IA was clearly

birthweight baby and spontaneous preterm, very recorded, 52 560 women were noted to have had surgical on September 24, 2021 by guest. Protected copyright. preterm and extremely preterm birth were significantly and 16 702, medical abortions. As table 5 shows, repro- higher following IA than in primigravid women. The risk ductive outcomes were comparable in the two groups of pre-eclampsia in women with a previous IA was higher except for a lower risk of a second IA following surgical than that in primigravid women but lower than that in termination of pregnancy. The adjusted RR of miscar- women with a previous miscarriage (table 2). riage, ectopic pregnancy, placenta praevia and sponta- The demographic characteristics of women who had neous preterm delivery (<37 weeks) were significantly a live birth in their first pregnancy and then went on to higher after surgical termination. In comparison with have an IA, a live birth or a miscarriage in their second primigravid women, that is, no previous abortion, pregnancy are shown in table 3. Women with an IA in women with a medical abortion had an increased risk of their second pregnancy were younger, belonged to placental abruption but not spontaneous preterm, very a more deprived social group and were more likely to be preterm or extremely preterm delivery. In contrast, smokers than women who had a live birth in their women with a surgical abortion had higher risks of all second pregnancy. Compared with women who had three types of spontaneous preterm delivery. They also a miscarriage in their second pregnancy, women with had an increased risk of pre-eclampsia, placenta praevia, a previous IA were older, belonged to more deprived placental abruption and low birthweight babies. More social classes and were more likely to smoke. women had repeat abortion following surgical termina- As table 4 shows, IA in the second pregnancy was tion of pregnancy, and fewer went on to have a live birth associated with a higher risk of an ectopic pregnancy or in comparison with primigravid women and those who an IA in the third pregnancy compared with an initial had medical terminations.

4 Bhattacharya S, Lowit A, Bhattacharya S, et al. BMJ Open 2012;2:e000911. doi:10.1136/bmjopen-2012-000911 htahraS oi ,Batcay S, Bhattacharya A, Lowit S, Bhattacharya

Table 2 Reproductive and perinatal outcomes following induced abortion, miscarriage or live birth in ﬁrst pregnancy Outcome in ﬁrst pregnancy Crude and adjusted (Adj.) RR (99% CI)* Induced abortion Induced abortion versus Induced abortion Live birth Miscarriage versus live birth miscarriage Outcome of second N¼120 033 N¼457 477 N¼47 355 pregnancy Live birth 67 336 (56.1) 355 674 (77.7) 36 479 (77.0) Crude 0.72 (0.71 to 0.73) Crude 0.72 (0.72 to 0.73) Adj. 0.74 (0.73 to 0.74) Adj. 0.69 (0.69 to 0.70) Stillbirth 409 (0.34) 1406 (0.31) 247 (0.52) Crude 1.11 (0.96 to 1.28) Crude 0.65 (0.53 to 0.80) tal et Adj. 1.06 (0.91 to 1.24) Adj. 0.58 (0.46 to 0.74) .

M Open BMJ Miscarriage 7965 (6.6) 30 669 (6.7) 6197 (13.1) Crude 0.99 (0.96 to 1.02) Crude 0.51 (0.49 to 0.53) Adj. 1.05(1.01 to 1.08) Adj. 0.56 (0.54 to 0.59) Ectopic pregnancy 1115 (0.9) 2939 (0.6) 499 (1.1) Crude 1.45 (1.32 to 1.58) Crude 0.88 (0.77 to 1.01) Adj. 1.36 (1.23 to 1.50) Adj. 0.83 (0.71 to 0.97) 2012; Induced abortion 43 208 (36.0) 66 789 (14.6) 3933 (8.3) Crude 2.47 (2.43 to 2.50) Crude 4.33 (4.16 to 4.51)

2 Adj. 2.30 (2.27 to 2.33) Adj. 4.64 (4.44 to 4.85) e091 doi:10.1136/bmjopen-2012-000911 :e000911. Outcome in ﬁrst pregnancy Crude and adjusted (Adj.) RR (99% CI)*

Induced Induced abortion Induced abortion versus Induced abortion abortion induced following outcomes Reproductive abortion Live birth Miscarriage Primigravida versus live birth miscarriage versus primigravida Outcomes in ongoing N¼67 745 N¼357 080 N¼36 726 N¼457 477 pregnancies Pre-eclampsia 1583 (2.3) 2982 (0.8) 922 (2.5) 8649 (1.9) Crude 2.80 (2.58 to 3.03) Crude 0.93 (0.84 to 1.03) Crude 1.24 (1.15 to 1.32) Adj. 2.42 (2.21 to 2.65) Adj. 0.83 (0.73 to 0.94) Adj. 1.26 (1.17 to 1.35) Placenta praevia 385 (0.6) 1919 (0.5) 289 (0.8) 2042 (0.5) Crude 1.06 (0.92 to 1.22) Crude 0.72 (0.59 to 0.88) Crude 1.27 (1.10 to 1.47) Adj. 1.09 (0.93 to 1.28) Adj. 0.79 (0.62 to 1.01) Adj. 1.05 (0.91 to 1.22) Abruptio placenta 339 (0.5) 1197 (0.3) 173 (0.5) 1770 (0.4) Crude 1.49 (1.27 to 1.75) Crude 1.06 (0.84 to 1.35) Crude 1.30 (1.11 to 1.51) Adj. 1.49 (1.25 to 1.77) Adj. 1.00 (0.76 to 1.32) Adj. 1.28 (1.10 to 1.50) Induction of laboury 18 044 (26.6) 69 482 (19.5) 10 347 (28.2) 120 080 (26.3) Crude 1.37 (1.34 to 1.39) Crude 0.95 (0.92 to 0.97) Crude 1.01 (1.00 to 1.03) Adj. 1.33 (1.30 to 1.35) Adj. 0.98 (0.95 to 1.01) Adj. 1.00 (0.99 to 1.02) Low birth weight 5385 (8.0) 16 309 (4.6) 3101 (8.5) 28 735 (6.3) Crude 1.74 (1.67 to 1.81) Crude 0.94 (0.89 to 1.00) Crude 1.27 (1.22 to 1.31) <2500 gz Adj. 1.24 (1.17 to 1.31) Adj. 0.96 (0.90 to 1.03) Adj. 1.08 (1.04 to 1.13) Outcomes in N¼45 656 N¼255 220 N¼23 751 N¼318 217 spontaneous births Spontaneous preterm 4224 (9.3) 13 453 (5.3) 2376 (10.0) 21 891 (6.9) Crude 1.76 (1.68 to 1.83) Crude 0.92 (0.86 to 0.97) Crude 1.35 (1.29 to 1.40) birth <37 weeks Adj. 1.66 (1.58 to 1.74) Adj. 0.85 (0.79 to 0.91) Adj. 1.37 (1.32 to 1.42) Spontaneous very 878 (1.9) 2157 (0.9) 513 (2.2) 4051 (1.3) Crude 2.28 (2.05 to 2.52) Crude 0.89 (0.77 to 1.03) Crude 1.51 (1.37 to 1.66) preterm birth <32 weeks Adj. 2.20 (1.96 to 2.47) Adj. 0.83 (0.70 to 0.99) Adj. 1.57 (1.43 to 1.72) Values are n (%) unless otherwise specified. Statistically significant relative risks are shown as bold. *Adjusted for maternal age, year of delivery, Carstairs at first pregnancy and interpregnancy interval. yFurther adjusted for pre-eclampsia, placenta praevia and abruptio placenta. zLow birth weight also adjusted for gestational age.

BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from from Downloaded 2012. August 6 on 10.1136/bmjopen-2012-000911 as published first Open: BMJ http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected by copyright. by Protected guest. by 2021 24, September on Reproductive outcomes following induced abortion BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from

Table 3 Demographic characteristics of women who had induced abortion, live birth or miscarriage after an initial live birth Outcome in second pregnancy following an initial live birth Induced abortion Live birth Miscarriage (N[30 527) (N[125 855) p Value (N[22 404) p Value Mean age (SD) 26.04 (5.85) 26.15 (4.68) <0.001 28.41 (5.42) 0.001 Carstairs category* y 1 3523 (12.8) 20 264 (16.5) <0.001 4498 (20.9) <0.001 2 4304 (15.6) 21 985 (17.9) 4079 (18.9) 3 5186 (18.8) 23 425 (19.0) 4312 (20.0) 4 6243 (22.6) 25 979 (21.1) 4447 (20.6) 5 8370 (30.3) 31 395 (25.5) 4235 (19.6) Smoking statusy Never 393 (39.7) 32 464 (48.5) <0.001 3165 (46.1) 0.001 Current 313 (31.6) 20 658 (30.9) 1169 (17.0) Former 43 (4.3) 5359 (8.0) 282 (4.1) Not known 241 (24.3) 8482 (12.7) 2243 (32.7) Total 990 66 963 6859 Missing 29 537 (96.8) 58 892 (46.8) 15 545 (69.4) Interpregnancy interval Median (IQR) 108 (61e209) 152 (96e256) <0.001 60 (48e87) <0.001 Values are n (%) unless otherwise speciﬁed. *Carstairs categories: 1¼ least deprived, 5¼ most deprived. yPercentage based on available information for each group.

Table 6 summarises the risk of spontaneous preterm utive IAs were not at signiﬁcantly higher risk of sponta- delivery in subsequent pregnancies following one or neous preterm birth in comparison with women who more consecutive IAs in comparison with those with no have had one termination of pregnancy. previous abortions (primigravid women). The adjusted RRs of spontaneous preterm birth (<37 weeks) was Strengths incrementally higher in women undergoing one, two, To our knowledge, this is the largest population-based

three and four IAs. The adjusted RRs of spontaneous study of reproductive outcomes following an IA. http://bmjopen.bmj.com/ very preterm delivery (<32 weeks) was higher after one Registry-based previous studies reporting preterm birth and four IAs, while the adjusted RRs of spontaneous rates as an outcome have been unable to discriminate extremely preterm delivery (<28 weeks) was higher between spontaneous and induced preterm delivery; this following two and four previous IAs. Additional IAs were is one of the ﬁrst papers to be able to calculate and not associated with increased adjusted RRs of any type of report spontaneous preterm birth rates after IA. spontaneous preterm birth. We have acknowledged changes in clinical practice over the years during which data were collected and have

adjusted for year of pregnancy in the regression models. on September 24, 2021 by guest. Protected copyright. DISCUSSION The choice of an appropriate comparison group to Principal findings women with a history of IA is problematic. Women who Our results suggest that women who had an IA in the are pregnant again after having undergone an IA in first pregnancy were more at risk of maternal and peri- a previous (first) pregnancy are gravida 2 and parity 0. It natal risks in comparison with women with a previous is impossible to control for both gravidity and parity live birth. Compared with an initial miscarriage, an IA in unless the unexposed cohort have had a prior preg- a first pregnancy was associated with a higher subsequent nancy, which did not lead to a delivery. Other compar- risk of miscarriage or ectopic pregnancy, IA and pre- ison groups can be either women in their first ongoing eclampsia. Women with a previous IA face increased risks pregnancies (gravidity 1 parity 0) or in their second of antepartum haemorrhage and spontaneous preterm ongoing pregnancies after a previous delivery (gravidity birth than women in their first pregnancy. 2 parity 1). We feel that our strategy comparing the A live birth prior to an IA does not appear to be exposed cohort with all three of the above groups adds associated with reduced perinatal complications in validity to our results. women who are at higher risk of spontaneous preterm birth than primigravida. Surgical termination appears to Limitations be associated with a higher chance of spontaneous The main limitations of this study stem from unrecorded preterm birth than medical IA. There does not appear to and missing data in relation to certain potential be a dose-dependent effect of IA on future adverse confounding factors within the data set. For example, perinatal outcomes. Women with three or four consec- smoking data were only available for 50% of women;

6 Bhattacharya S, Lowit A, Bhattacharya S, et al. BMJ Open 2012;2:e000911. doi:10.1136/bmjopen-2012-000911 htahraS oi ,Batcay S, Bhattacharya A, Lowit S, Bhattacharya

Table 4 Reproductive and perinatal outcomes in women who had induced abortion, live birth or miscarriage following a live birth in the ﬁrst pregnancy

Outcome of second pregnancy Induced abortion Induced abortion versus Induced abortion Live birth Miscarriage versus live birth miscarriage Outcome of third pregnancy N¼30 527 N¼125 855 N¼22 404 Crude and adjusted (Adj.) RR (99% CI)* Live birth 18 562 (60.8) 85 014 (67.5) 17 745 (79.2) Crude 0.90 (0.89 to 0.91) Crude 0.77 (0.76 to 0.78) tal et Adj. 0.88 (0.87 to 0.89) Adj. 0.77 (0.76 to 0.78)

. Stillbirth 84 (0.3) 426 (0.3) 69 (0.3) Crude 0.81 (0.60 to 1.11) Crude 0.89 (0.59 to 1.36) M Open BMJ Adj. 0.76 (0.55 to 1.06) Adj. 0.86 (0.54 to 1.37) Miscarriage 2005 (6.6) 8778 (7.0) 2869 (12.8) Crude 0.94 (0.89 to 1.00) Crude 0.51 (0.48 to 0.55) Adj. 0.93 (0.88 to 1.00) Adj. 0.67 (0.62 to 0.72) 2012; Ectopic pregnancy 339 (1.1) 1064 (0.9) 181 (0.8) Crude 1.31 (1.12 to 1.54) Crude 1.38 (1.09 to 1.74) Adj. 1.31 (1.11 to 1.56) Adj. 1.16 (0.90 to 1.50) 2 e091 doi:10.1136/bmjopen-2012-000911 :e000911. Induced abortion 9537 (31.2) 30 573 (24.3) 1540 (6.9) Crude 1.29 (1.25 to 1.32) Crude 4.55 (4.25 to 4.86) Adj. 1.33 (1.30 to 1.37) Adj. 4.37 (4.06 to 4.70) Outcomes in ongoing pregnancies N¼18 646 N¼85 440 N¼17 814 Crude and adjusted (Adj.) RR (99% CI)y abortion induced following outcomes Reproductive Pre-eclampsia 144 (0.8) 567 (0.7) 165 (0.9) Crude 1.16 (0.92 to 1.48) Crude 0.83 (0.62 to 1.12) Adj. 1.40 (1.10 to 1.79) Adj. 0.91 (0.66 to 1.27) Placenta praevia 183 (1.0) 473 (0.6) 133 (0.8) Crude 1.77 (1.42 to 2.22) Crude 1.32 (0.98 to 1.76) Adj. 1.78 (1.40 to 2.25) Adj. 1.34 (0.97 to 1.84) Abruptio placenta 91 (0.5) 325 (0.4) 66 (0.4) Crude 1.28 (0.95 to 1.74) Crude 1.32 (0.87 to 2.00) Adj. 1.28 (0.93 to 1.77) Adj. 1.32 (0.83 to 2.10) Induction of labourz 4298 (23.1) 18 239 (21.4) 3968 (22.3) Crude 1.08 (1.04 to 1.12) Crude 1.03 (0.98 to 1.09) Adj. 1.11 (1.07 to 1.16) Adj. 1.01 (0.96 to 1.07) Low birth weight <2500x 1086 (5.8) 3905 (4.6) 784 (4.4) Crude 1.28 (1.17 to 1.39) Crude 1.32 (1.17 to 1.49) Adj. 1.36 (1.21 to 1.51) Adj. 1.04 (0.90 to 1.21) Outcomes in spontaneous births N¼12 868 N¼59 220 N¼12 056 Spontaneous preterm 859 (6.7) 3035 (5.1) 644 (5.3) Crude 1.30 (1.18 to 1.43) Crude 1.25 (1.10 to 1.42) birth <37 weeks Adj. 1.27 (1.14 to 1.40) Adj. 1.14 (0.99 to 1.32) Spontaneous very preterm birth <32 weeks 162 (1.3) 495 (0.8) 104 (0.9) Crude 1.51 (1.19 to 1.90) Crude 1.46 (1.06 to 2.01) Adj. 1.44 (1.12 to 1.84) Adj. 1.35 (0.93 to 1.96) Values are n (%) unless otherwise speciﬁed. Statistically signiﬁcant relative risks are shown in bold. *Adjusted for age, year of delivery, Carstairs at second pregnancy and interpregnancy interval. yAdjusted for maternal age, year of pregnancy, Carstairs category at second pregnancy and interpregnancy interval. zFurther adjusted for pre-eclampsia, placenta praevia and abruptio placenta. xLow birth weight also adjusted for gestational age.

BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from from Downloaded 2012. August 6 on 10.1136/bmjopen-2012-000911 as published first Open: BMJ http://bmjopen.bmj.com/ on September 24, 2021 by guest. Protected by copyright. by Protected guest. by 2021 24, September on 8 erdcieotoe olwn nue abortion induced following outcomes Reproductive

Table 5 Reproductive outcomes following medical and surgical abortion Medical vs Surgical vs Surgical vs medical Reproductive Medical termination Surgical termination Primigravida crude Primigravida crude induced abortion outcomes in next Primigravida in ﬁrst pregnancy in ﬁrst pregnancy and adjusted (Adj.) and adjusted (Adj.) crude and adjusted (2nd) pregnancy N[457 477 N[16 702 N[52 560 RR RR (99% CI)* RR (99% CI)* (Adj.) RR (99% CI)* Live birth 355 674 (77.7) 9785 (58.6) 28 285 (53.8) Crude Adj. 0.75 (0.74 to 0.77) 0.69 (0.69 to 0.70) 0.92 (0.90 to 0.94) 0.71 (0.70 to 0.73) 0.76 (0.75 to 0.77) 1.44 (1.41 to 1.48) Still birth 1406 (0.3) 57 (0.3) 151 (0.3) Crude Adj. 1.11 (0.79 to 1.57) 0.93 (0.75 to 1.17) 0.84 (0.56 to 1.26) htahraS oi ,Batcay S, Bhattacharya A, Lowit S, Bhattacharya 1.15 (0.80 to 1.64) 0.95 (0.73 to 1.23) 0.98 (0.57 to 1.69) Miscarriage 30 669 (6.7) 1200 (7.2) 3723 (7.1) Crude Adj. 1.07 (1.00 to 1.15) 1.06 (1.01 to 1.10) 0.99 (0.91 to 1.07) 0.98 (0.91 to 1.06) 1.03 (0.98 to 1.08) 1.45 (1.30 to 1.62) Ectopic 2939 (0.6) 120 (0.7) 599 (1.1) Crude Adj. 1.12 (0.88 to 1.42) 1.77 (1.58 to 1.99) 1.59 (1.23 to 2.05) 0.99 (0.78 to 1.28) 1.80 (1.58 to 2.06) 1.78 (1.29 to 2.45) Induced abortion 66 789 (14.6) 5540 (33.2) 19 802 (37.7) Crude Adj. 2.27 (2.21 to 2.34) 2.58 (2.54 to 2.63) 1.14 (1.10 to 1.17) 3.01 (2.91 to 3.12) 2.00 (1.96 to 2.04) 0.44 (0.42 to 0.46) Outcome in ongoing N¼457 477 N¼9842 N¼28 436 pregnancyx Pre-eclampsia 8649 (1.9) 316 (3.2) 688 (2.4) Crude Adj. 1.70 (1.47 to 1.96) 1.28 (1.16 to 1.42) 0.75 (0.63 to 0.90) 1.01 (0.86 to 1.17) 1.14 (1.03 to 1.27) 1.12 (0.90 to 1.39) Placenta praevia 2042 (0.5) 23 (0.2) 248 (0.9) Crude Adj. 0.52 (0.31 to 0.90) 1.95 (1.64 to 2.32) 3.73 (2.13 to 6.54) tal et 0.81 (0.47 to 1.40) 1.63 (1.36 to 1.95) 2.23 (1.17 to 4.26)

. Abruptio placentae 1770 (0.4) 40 (0.4) 160 (0.6) Crude Adj. 1.05 (0.70 to 1.58) 1.45 (1.18 to 1.80) 1.38 (0.88 to 2.18) M Open BMJ 1.65 (1.08 to 2.52) 1.54 (1.24 to 1.91) 1.09 (0.63 to 1.88) Birth weighty <2500 g 28 735 (6.3) 697 (7.1) 2407 (8.5) Crude Adj. 1.13 (1.03 to 1.24) 1.35 (1.28 to 1.42) 1.19 (1.07 to 1.33) 1.05 (0.94 to 1.17) 1.16 (1.08 to 1.23) 1.12 (0.97 to 1.28)

2012; Spontaneous birthsx N[318 217z N[6474z N[18 126z Preterm <37 weeks 21891 (6.9) 533 (8.2) 1768 (9.8) Crude Adj. 1.20 (1.07 to 1.33) 1.42 (1.34 to 1.51) 1.18 (1.05 to 1.34) 2 e091 doi:10.1136/bmjopen-2012-000911 :e000911. 1.11 (0.99 to 1.24) 1.45 (1.37 to 1.55) 1.25 (1.07 to 1.45) Very preterm <32 weeks 4051 (1.3) 123 (1.9) 363 (2.0) Crude Adj. 1.49 (1.18 to 1.89) 1.57 (1.37 to 1.81) 1.05 (0.81 to 1.38) 1.25 (0.98 to 1.60) 1.62 (1.41 to 1.87) 1.13 (0.81 to 1.58) Very preterm <28 weeks 1349 (0.4) 35 (0.5) 120 (0.7) Crude Adj. 1.27 (0.82 to 2.00) 1.56 (1.22 to 2.00) 1.23 (0.75 to 2.01) 0.91 (0.58 to 1.44) 1.62 (1.27 to 2.07) 1.38 (0.73 to 2.61) Values are n (%) unless otherwise speciﬁed. Statistically signiﬁcant relative risks are shown as bold. *All relative risks comparing medical vs surgical have been adjusted for maternal age, year of event, Carstairs category at the previous & interpregnancy interval. yLow birth weight also adjusted for gestational age. zOnly spontaneous delivery considered among live & still birth.

xAll relative risks comparing primigravida vs medical/surgical have been adjusted for maternal age, year of event, Carstairs category at the ongoing pregnancy.

data on body mass index were unavailable, while data on gestational age at termination was missing in the majority of cases. The actual method of termination (medical vs surgical) was unrecorded in around 25% of all cases, while a large number of women appeared to have both medical and surgical treatment. Parity number was less reliable in the early years of data 1.12 (0.18 to 6.96) 5.96 (1.65 to 21.37) 1.89 (1.08 to 3.31) collection. This may reﬂect problems with coding and could potentially affect the quality of our results. In addition, the analysis of such a large population-based data set has the capacity to produce statistically signiﬁ- cant differences, which may or may not be clinically relevant, although this has been minimised by our use of 28 weeks

< a stringent 1% signiﬁcance level throughout. Deﬁning an ideal reference group is a challenge in 6.94 (1.95 to 24.72) 1.44 (1.09 to 1.87) 1.24 (0.94 to 1.64) PTD studies exploring outcomes after IA. While we have partially addressed this issue by using more than one unexposed cohort, our data do not allow us to adjust for potential differences in pregnancy intentions between groups, which can impact on antenatal care and perinatal outcomes. Unrecorded data relating to key potential confounders cannot exclude the possibility that some 4.27 (1.76 to 10.37) 1.59 (1.37 to 1.84) associations are not explained by abortion itself but by special circumstances of women seeking abortion, which also increases their risk of complications in pregnancy. We ran a separate analysis to identify previous pregnancy complications in women who had an IA, a miscarriage or a live birth in their second pregnancy. As supplementary

32 weeks table A shows, IA in the second pregnancy was not < signiﬁcantly associated with increased RR (99% CI) of http://bmjopen.bmj.com/ PTD 4.62 (1.91 to 11.19) 1.81 (0.74 to 4.46) 1.64 (0.67 to 4.06) 1.36 (0.22 to 8.37) 1.70 (1.47 to 1.96) 1.48 (1.00 to 2.19) 1.34 (0.91 to 2.00) 2.27 (1.31 to 3.94) pre-eclampsia, placenta praevia, placental abruption and low birth weight, respectively, compared with live birth (0.99 (0.85 to 1.16), 1.29 (0.99 to 1.67), 1.32 (0.96 to 1.82) and 1.08 (0.98, 1.18)) or miscarriage (0.79 (0.65 to 0.96), 1.17 (0.81 to 1.69), 1.08 (0.70 to 1.68) and 1.14 (1.00 to 1.30)). Comparison with previous studies

The association between IA and preterm birth found in on September 24, 2021 by guest. Protected copyright. 2.10 (1.23 to 3.59) 1.52 (1.01 to 2.27) 1.47 (1.38 to 1.57) 1.51 (1.29 to 1.77) this study is consistent with previously published work.32 Two recent meta-analyses suggest that women who have had an IA are at higher risk of preterm birth in subsequent pregnancies.33 34 Our study shows that after adjustment women with a previous abortion have an weeks increased chance of a subsequent preterm birth and very 37 < preterm birth compared with primigravidae or those who have had a previous live birth, but at no signiﬁcantly PTD Crude RR (99% CI) Adj.RR (99% CI)* Crude RR (99% CI) Adj RR (99% CI) Crude RR (99% CI) Adj RR (99% CI) greater risk compared with women who have had a previous miscarriage. Women who had a live birth before an IA are more likely to have a preterm birth compared with women with two previous live births. Our results did not suggest a signiﬁcant increased risk of miscarriage after an IA, which is in keeping with 21 35

Risk of spontaneous preterm delivery following increasing number of induced abortions a review of literature. In contrast, Sun et al (2003) demonstrated an association between surgical abortion and miscarriage in a subsequent pregnancy. Literature on the association between IA and miscarriage or ectopic 4 Previous abortions vs 02 Previous abortions 2.13 vs (1.263 1 to Previous 3.64) abortions vs4 2 Previous abortions 1.03 vs (0.87 3 to 1.22) 1.00 (0.65 to 1.54) 1.38 (0.71 to 2.68) 1.02 (0.86 to 1.21) 1.01 (0.66 to 1.55) 1.38 (0.71 0.87 to (0.57 2.70) to 1.31) 1.23 (0.46 to 3.27) 2.55 (0.72 to 9.01) 0.84 (0.56 to 1.28) 1.22 (0.46 to 3.26) 2.60 (0.74 to 9.18) 1.58 (0.86 to 2.89) 0.60 (0.09 to 3.99) 5.10 (0.56 to 46.78) 1.52 (0.83 to 2.78) 0.60 (0.09 to 5.29 (0.58 3.97) to 48.70) 3 Previous abortions vs 0 1.55 (1.04 to 2.31) 1 Previous abortion vs2 0 Previous abortions vs 0 1.50 (1.41 to 1.55 1.59) (1.32 to 1.81) Table 6 Statistically significant relative risksPTD, are preterm shown delivery. as*Adjusted bold. for maternal age, year of delivery, Carstairs at first pregnancy and interpregnancy interval. pregnancy is sparse and conflicting. The increased risk

Bhattacharya S, Lowit A, Bhattacharya S, et al. BMJ Open 2012;2:e000911. doi:10.1136/bmjopen-2012-000911 9 Reproductive outcomes following induced abortion BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from of having a second termination following IA in a first first trimester, medical IAs. While this could reflect the pregnancy highlighted in our study has been reported effect of repeated surgical trauma to the cervix, e elsewhere.36 38 While women who had an abortion were this needs further exploration in future studies with more likely to have a subsequent abortion, but they may long-term periods of follow-up. also be more likely to have an unintended pregnancy. A key challenge in studying health sequelae after IA is This should be seen a potential risk factor, which should to deal with potential differences in pregnancy inten- be explored in future studies. tions between comparison groups. While women who Available literature suggests that there is an association had an abortion were more likely to have a subsequent between IA and placenta praevia,39 40 but no association abortion, they may also be more likely to have an unin- with abruptio placenta.41 42 This study found that women tended pregnancy, which needs to be acknowledged as in their second pregnancy after an initial IA in the first a potential risk factor in future studies. were at higher odds of both placenta praevia and abruptio placenta; women in their third pregnancy after an IA in CONCLUSIONS their second pregnancy had higher odds of placenta IA in a first pregnancy is associated with a higher risk of praevia but not abruptio placenta. Published evidence spontaneous preterm birth in a subsequent pregnancy in 43 44 supports a decreased risk of pre-eclampsia after an IA. comparison with primigravid women but not in women Our results suggest a risk of developing pre-eclampsia, with a previous miscarriage. A successful pregnancy which is on par with primigravid women, but lower than leading to a live birth prior to an IA does not appear to women with a previous miscarriage. The reasons for these ameliorate this risk, while more than one abortion does associations are unclear, and hence any explanations can not significantly increase it. Surgical, but not medical only be speculative. Problems with placental position and abortion appears to be associated with an increased risk function could occur due to disruption of the endome- of spontaneous very preterm birth in comparison with trium by vigorous curettage. The quality of placental primigravid women. The results of this study should help function in a previous pregnancy could influence provide women as well as health professionals with susceptibility to future pre-eclampsia. accurate information to inform clinical decision-making Since the introduction of medical abortion, there has and tailor antenatal care to address women’s risk been much speculation about the rival merits of medical profiles. and surgical techniques, especially in terms of future reproductive outcomes. Analysis of Danish data has Acknowledgements We thank staff at ISD Scotland for extraction of data failed to demonstrate a difference in key outcomes, such from the Scottish Morbidity Records Database and Margery Heath for secretarial assistance. as preterm birth between medical and surgical abortion, but this study was unable to identify spontaneous versus Contributors AT conceived the idea for the study. SB was the principal http://bmjopen.bmj.com/ 30 investigator. He designed the study along with SohB, AT, AJL and TM, led the induced preterm birth. With our ability to identify funding application, managed the project, interpreted the results and wrote the spontaneous preterm births we have shown a clear first draft of the paper. AL cleaned the data and performed some of the initial association with surgical abortion. However, since we analyses. SohB cowrote the funding application, facilitated data manipulation, were unable to adjust for gestational age, we cannot rule interpreted the results and helped to draft the paper. EAR performed the out the possibility that surgical abortions may have been statistical analysis and interpreted the results with input from AJL. All authors commented on, and contributed to the final draft of the paper. performed at a more advanced stage of pregnancy requiring a greater degree of cervical dilatation, thus Funding The Chief Scientist Office Scotland funded the study (grant number CZG/2/403). The views expressed are those of the authors and not the leading to future preterm labour. Our results are on September 24, 2021 by guest. Protected copyright. funding body. supported by a recent publication showing that the risk of preterm birth after one or more surgical abortions is Competing interests The authors declare that they have no competing interests. higher than after medical abortion and comparable to that in primigravid women.11 Ethics approval The ethics approval was provided by the Privacy Advisory A dose-dependent relationship between the number of Committee of Information and Services Division, NHS, Scotland. IAs and future PTB has been shown in a number of Provenance and peer review Not commissioned; externally peer reviewed. 32 previous studies. The results of our analysis do not Data sharing statement There are no unpublished data available for support this. Given our inability to adjust for a number confidentiality reasons. of potential confounders, this needs to be investigated further. REFERENCES Our data suggest that medical and surgical termina- 1. http://www.isdscotland.org/isd/CCC_FirstPage.jsp- 2. Berkowitz GS, Papiernik E. Epidemiology of preterm birth. Epidemiol tions may impact differently on future reproductive Rev 1993;15:414e43. outcomes, with a higher risk of spontaneous preterm 3. Calhoun BC, Shadigian E, Rooney B. Cost consequences of induced birth after surgery. We were unable to disentangle abortion as an attributable risk for preterm birth and impact on informed consent. J Reprod Med 2007;52:929e37. the separate effects of repeated medical and surgical 4. Chen A, Yuan W, Meirik O, et al. Mifepristone-induced early abortion abortion due to a relative paucity of numbers. and outcome of subsequent wanted pregnancy. Am J Epidemiol 11 2004;160:110e17. A recent publication found an increased risk of 5. de Haas I, Harlow BL, Cramer DW, et al. Spontaneous preterm birth: premature delivery following multiple surgical, but not a case-control study. Am J Obstet Gynecol 1991;165:1290e6.

10 Bhattacharya S, Lowit A, Bhattacharya S, et al. BMJ Open 2012;2:e000911. doi:10.1136/bmjopen-2012-000911 Reproductive outcomes following induced abortion BMJ Open: first published as 10.1136/bmjopen-2012-000911 on 6 August 2012. Downloaded from

6. Foix-L’He´lias L, Blondel B. Changes in risk factors of preterm delivery 24. Moreau C, Kaminski M, Ancel PY, et al. Previous induced abortions in France between 1981 and 1995. Paediatr Perinat Epidemiol and the risk of very preterm delivery: results of the EPIPAGE study. 2000;14:314e23. BJOG 2005;112:430e7. 7. Freak-Poli R, Chan A, Tucker G, et al. Previous abortion and risk of 25. Zhou W, Nielsen GL, Larsen H, et al. Induced abortion and placenta pre-term birth: a population study. J Matern Fetal Neonatal Med complications in the subsequent pregnancy. Acta Obstet Gynecol 2009;22:1e7. Scand 2001;80:1115e20. 8. Goldenberg RL, Culhane JF, Iams JD, et al. Epidemiology and 26. Henriet L, Kaminski M. Impact of induced abortions on subsequent causes of preterm birth. Lancet 2008;371:75e84. pregnancy outcome: the 1995 French national perinatal survey. 9. Lang JM, Lieberman E, Cohen A. A comparison of risk factors for BJOG 2001;108:1036e42. preterm labor and term small-for-gestational-age birth. Epidemiology 27. Henshaw RC, Naji SA, Russell IT, et al. A comparison of medical 1996;7:369e76. abortion (using mifepristone and gemeprost) with surgical vacuum 10. Lekea-Karanika V, Tzoumaka-Bakoula C, Golding J. Previous aspiration: efficacy and early medical sequelae. Hum Reprod obstetric history and subsequent preterm delivery in Greece. Eur J 1994;9:2167e72. Obstet Gynecol Reprod Biol 1990;37:99e109. 28. Ashok PW, Kidd A, Flett GM, et al. A randomized comparison of 11. Liao H, Wei Q, Duan L, et al. Repeated medical abortions and the risk medical abortion and surgical vacuum aspiration at 10-13 weeks of preterm birth in the subsequent pregnancy. Arch Gynecol Obstet gestation. Hum Reprod 2002;17:92e8. 2011;284:579e86. 29. Rørbye C, Nørgaard M, Nilas L. Medical versus surgical abortion 12. Martius JA, Steck T, Oehler MK, et al. Risk factors associated with efficacy, complications and leave of absence compared in a partly preterm (<37+0 weeks) and early preterm birth (<32+0 weeks): randomized study. Contraception 2004;70:393e9. univariate and multivariate analysis of 106,345 singleton births from 30. Virk J, Zhang J, Olsen J. Medical abortion and the risk of subsequent the 1994 statewide perinatal survey of Bavaria. Eur J Obstet Gynecol adverse pregnancy outcomes. N Engl J Med 2007;357:648e53. Reprod Biol 1998;80:184e9. 31. Carstairs V, Morris R. Deprivation and health in Scotland. Aberdeen: 13. Pickering RM, Deeks JJ. Risks of delivery during the 20th to the 36th Aberdeen University Press 1991. week of gestation. Int J Epidemiol 1991;20:456e66. 32. Lowit A, Bhattacharya S. Obstetric performance following an induced 14. Reime B, Schu¨cking BA, Wenzlaff P. Reproductive outcomes in abortion. Best Pract Res Clin Obstet Gynaecol 2010;24:667e82. adolescents who had a previous birth or an induced abortion 33. Shah PS, Zao J. Induced termination of pregnancy and low birth compared to adolescents’ first pregnancies. BMC Pregnancy Childbirth weight and preterm birth: a systemic review and meta-analyses. 2008;8:4. BJOG 2009;116:1425e42. 15. Voigt M, Olbertz D, Fusch C, et al. The influence of previous 34. Swingle HM, Colaizy TT, Zimmerman MB, et al. Abortion and the risk pregnancy terminations, miscarriages and still-births on the incidence of subsequent preterm birth: a systematic review with meta-analyses. of babies with low birth weight and premature births as well as J Reprod Med 2009;54:95e108. a somatic classification of newborns (In German). Z Geburtshilfe 35. Sun Y, Che Y, Gao E, et al. Induced abortion and risk of subsequent Neonatol 2008;212:5e12. miscarriage. Int J Epidemiol 2003;32:449e54. 16. Watson LF, Rayner JA, King J, et al. Modelling sequence of prior 36. Heikinheimo O, Gissler M, Suhonen S. Age, parity, history of abortion pregnancies on subsequent risk of very preterm birth. Paediatr and contraceptive choices affect the risk of repeat abortion. Perinat Epidemiol 2010;24:416e23. Contraception 2008;78:150e4. 17. Winer N, Resche-Rigon M, Morin C, et al. Is induced abortion with 37. Prager SW, Steinauer JE, Foster DF, et al. Risk factors for repeat misoprostol a risk factor for late abortion or preterm delivery in elective abortion. American Journal of Obstetrics & Gynecology subsequent pregnancies? Eur J Obstet Gynecol Reprod Biol 2007;197:575.e1e6. 2009;145:53e6. 38. Rowlands S. More than one abortion. J Fam Plann Reprod Health 18. Zhou W, Sørensen HT, Olsen J. Induced abortion and subsequent Care 2007;33:155e8. pregnancy duration. Obstet Gynecol 1999;94:948e53. 39. Ananth CV, Smulian JC, Vintzileos AM. The effect of placenta previa 19. Atrash HK, Hogue CJ. The effect of pregnancy termination on future on neonatal mortality: a population-based study in the United States, reproduction. Baillieres Clin Obstet Gynaecol 1990;4:391e405. 1989 through 1997. Am J Obstet Gynecol 2003;188:1299e304. 20. Hogue CJ. Impact of abortion on subsequent fecundity. Clin Obstet 40. Hung TH, Hsieh CC, Hsu JJ, et al. Risk factors for placenta abruption http://bmjopen.bmj.com/ Gynaecol 1986;13:96e103. in an Asian population. Reprod Sci 2007;14:59e65. 21. Thorp JM Jr, Hartmann KE, Shadigian E. Long-term physical and 41. Hung TH, Hsieh CC, Hsu JJ, et al. Risk factors for placenta previa in psychological health consequences of induced abortion: review of the an Asian population. Int J Gynaecol Obstet 2007;97:26e30. evidence. Obstet Gynecol Surv 2003;58:67e79. 42. Zhu QX, Gao ES, Chen AM, et al. Mifepristone-induced abortion and 22. Ancel PY, Lelong N, Papiernik E, et al; EUROPOP. History of placenta complications in subsequent pregnancy. Hum Reprod induced abortion as a risk factor for preterm birth in European 2009;24:315e19. countries: results from the EUROPOP survey. Hum Reprod 43. Eras JL, Saftlas AF, Triche E, et al. Abortion and its effect on risk of 2004;19:734e40. preeclampsia and transient hypertension. Epidemiology 23. Raatikainen K, Heiskanen N, Heinonen S. Induced abortion: not an 2000;11:36e43. independent risk factor for pregnancy outcome, but a challenge for 44. Trogstad L, Magnus P, Skjaerven R, et al. Previous abortions and risk

health counseling. Ann Epidemiol 2006;16:587e92. of pre-eclampsia. Int J Epidemiol 2008;37:1333e40. on September 24, 2021 by guest. Protected copyright.

PAGE fraction trail=10.75

Bhattacharya S, Lowit A, Bhattacharya S, et al. BMJ Open 2012;2:e000911. doi:10.1136/bmjopen-2012-000911 11